1. Field of the Invention
The present invention relates to an image processing technique for performing a variable-magnification process on an image.
2. Description of the Related Art
Hitherto, an image processing apparatus such as a digital color copying apparatus or the like has widely been known. According to such an image processing apparatus, light is irradiated to an original, its reflection light is read by a photoelectric converting device such as a CCD or the like, and various kinds of image processes such as a variable-magnification processing and the like can be executed to a read signal. As a method of irradiating the light to the original, it is possible to construct in such a manner that the original is fixed and an optical system is moved or the optical system is fixed and the original is conveyed. When the original is read by the image processing apparatus such as a digital color copying apparatus or the like, a linear image sensor in which photodiodes are arranged in a line and photoelectric converting devices such as CCDs or the like are arranged in parallel is generally used.
Hitherto, in the foregoing variable-magnification process, the magnification in the main scanning direction is realized by an image process by using the signal read by the linear image sensor. The magnification in the sub-scanning direction as a moving direction of the optical system is realized by scanning the reading optical system to the original at a low speed upon magnification or at a high speed upon reduction. According to such a method, within a control range of a motor to scan the optical system, since a driving speed can be finely adjusted, a relatively stable image signal can be obtained.
By using a buffer memory of a predetermined capacity, the magnification in the sub-scanning direction can be also realized by the image process in a manner similar to the magnification in the main scanning direction. In this case, a construction in which the signal read by the linear image sensor is temporarily stored into the buffer memory of the predetermined capacity, a CPU, dedicated hardware, or the like reads out the signal temporarily stored in the buffer memory, and the variable-magnification process is executed is generally used. According to such a construction, a scanning speed of the optical system can be held constant, control of a motor to scan the optical system can be simplified, and the reasonable motor can be used.
In association with the above technique, Japanese Patent Application Laid-Open No. H06-121116A discloses such a technique that a desired variable-magnification ratio in the sub-scanning direction is realized by combining a magnification in the sub-scanning direction which is performed by controlling a driving speed of a line sensor and a magnification in the main scanning direction which is performed by an image process. Japanese Patent Application Laid-Open No. 2003-259092A discloses such a technique that at a predetermined magnification or more, a scanning speed of an optical system is controlled, at a magnification less than the predetermined magnification, the scanning speed of the optical system is held constant, a photosensing time of a linear image sensor is controlled, and a magnification in the sub-scanning direction is performed.
However, in the case where the scanning speed of the optical system is held constant and the magnification by the image process is executed, if the magnification exceeds the predetermined variable-magnification ratio upon magnification, the buffer memory will overflow. For example, when the variable-magnification ratio is equal to 200%, a data amount is increased by four times after the variable-magnification process. Therefore, a communication speed which is permitted to the input side to a magnification processing unit prior to the variable-magnification process is equal to 1/4  of a communication speed at the time of a direct copying mode of the variable-magnification ratio of 100%. Therefore, a reading speed from the buffer memory is also equal to 1/4  of a reading speed at the time of the direct copying mode of the variable-magnification ratio of 100%. On the other hand, since the scanning speed of the optical system is held constant, a writing speed into the buffer memory cannot be reduced. Thus, the writing speed into the buffer memory is higher than the reading speed from the buffer memory and the buffer memory overflows.
Therefore, a method whereby the signal read by the linear image sensor is stored into the buffer memory of such a capacity that the overflow cannot occur and, thereafter, the variable-magnification process by the image process is executed is considered. The buffer memory of such a capacity that the overflow cannot occur is a ring buffer or a page memory in which RAW data is stored. However, according to such a method, the buffer memory of a large capacity is necessary and even if the data is multivalue compressed by a JPEG or the like and, thereafter, stored into the buffer memory, in the case of the whole one page, the data amount is sufficiently large and the buffer memory of the large capacity is necessary.
The invention is made to solve the foregoing problems and it is an object of the invention to provide an image processing apparatus in which even if a scanning speed of an optical system is held constant, a use amount of a buffer memory can be reduced.